Research map / Detection and segmentation / CNN features
Detecting Every Object From Events
Haitian Zhang, Chang Xu, Xinya Wang, Bingde Liu, Guang Hua, Lei Yu, Wen Tao Yang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add disentangled objectness head (ablation) | CA-RVT · Four Hour 1 Mpx, 5+2 | AUC (unknown classes): 17.77 -> 20.79 | Improved |
| add dual regressor head with potential samples (ablation) | CA-RVT · Four Hour 1 Mpx, 5+2 | AUC (unknown classes): 17.77 -> 20.29 | Improved |
| add both heads (ablation) | CA-RVT · Four Hour 1 Mpx, 5+2 | AUC (unknown classes): 17.77 -> 22.70 | Improved |
| add dual regressor head alone, measured on all classes (ablation) | CA-RVT · Four Hour 1 Mpx, 5+2 | AUC (all classes): 34.88 -> 33.83 | Hurt |
| add positive-only branch alongside positive-negative branch (ablation) | Disentangled head · Four Hour 1 Mpx, 5+2 | AR30 (unknown classes): 22.49 -> 26.00 | Improved |
| use positive-only branch alone instead of positive-negative (ablation) | Objectness head · Four Hour 1 Mpx, 5+2 | AR10 (all classes): 41.38 -> 33.79 | Mixed |
| raise potential sample count from 0 to 35 (ablation) | DEOE · Four Hour 1 Mpx, 5+2 | AUC (unknown classes): 20.25 -> 22.70 | Improved |
| raise potential sample count from 35 to 75 (ablation) | DEOE · Four Hour 1 Mpx, 5+2 | AUC (unknown classes): 22.70 -> 19.62 | Hurt |
| use DEOE instead of CA-RVT baseline | RVT · Four Hour 1 Mpx, 5+2 | AUC (unknown classes): 17.77 -> 22.70 | Improved |
| use DEOE instead of CA-RVT baseline | RVT · Four Hour 1 Mpx, 4+3 | AUC (unknown classes): 8.83 -> 12.49 | Improved |
| use DEOE instead of CA-RVT_P in cross-dataset test | RVT · DSEC-Detection (cross-dataset) | AUC (unknown classes): 11.16 -> 15.07 | Improved |
| use DEOE instead of CA-RVT, inference time | RVT · Four Hour 1 Mpx, 5+2 | Inference time (ms): 6.19 -> 8.82 | Hurt |
Closest papers by meaning
- LEOD: Label-Efficient Object Detection for Event Cameras (CVPR 2024 · similarity 0.88)
- FlexEvent: Towards Flexible Event-Frame Object Detection at Varying Operational Frequencies (NeurIPS 2025 · similarity 0.87)
- Ev-3DOD: Pushing the Temporal Boundaries of 3D Object Detection with Event Cameras (CVPR 2025 · similarity 0.87)
- Scene Adaptive Sparse Transformer for Event-based Object Detection (CVPR 2024 · similarity 0.86)
- EGSST: Event-based Graph Spatiotemporal Sensitive Transformer for Object Detection (NeurIPS 2024 · similarity 0.86)
- Object Detection using Event Camera: A MoE Heat Conduction based Detector and A New Benchmark Dataset (CVPR 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
